二、分析结果的统计学方法——显著性测验我们平时所搜集的资料,多数具有样本性质,比较这些平均数或相对数时,首先考虑到抽样误差问题。如果两个平均数或相对数不相同时,不能只从表面的差异(差别)作出结论。由于抽样误差所引起的差别,统计学上认为平均数或相对数的这种差异没有显著意义;若是这种差异由于同一个同质总体内的个体的变异性(不能用抽样误差来解的)而引起的,统计学认为这种差异是有显著意义的。 Second, the statistical methods to analyze the results - significance test We usually collect the data, the majority of the nature of the sample, the comparison of these average or relative, the first consideration of the sampling error. If the two averages or relative numbers are not the same, you can not conclude only from the surface differences (differences). Due to the difference caused by sampling error, there is no statistically significant difference between mean or relative numbers. If this difference is due to the variability of individuals in the same homogeneous population (can not be solved by sampling error) And caused, statistics that this difference is significant.